SMM-Conv computes convolutions by accumulating kernel-scaled shifted sub-matrices into the output, reusing a single buffer, and reports speedups over im2col+GEMM and MEC on CPU.
An updated set of basic linear algebra subprograms (blas)
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
SMM-Conv: Scalar Matrix Multiplication with Zero Packing for Accelerated Convolution
SMM-Conv computes convolutions by accumulating kernel-scaled shifted sub-matrices into the output, reusing a single buffer, and reports speedups over im2col+GEMM and MEC on CPU.